Integrity, Interrogation and Criminal Injustice

3 Integrity, Interrogation and Criminal Injustice

DAVID DIXON

[W]hat matters is how the confession was obtained, not whether or not it may have been true … [T]he rationale of these provisions is that they are intended, among other things, to preserve the integrity of the Common Law principle, that no person is required or can be made to incriminate himself. (Mitchell J, ruling on the voir dire in R v Heron , Leeds Crown Court, 1 November 1993)

Introduction: Two Murders, Two Flawed Investigations

This chapter explores the concept of integrity in criminal process through a close reading of the events surrounding two murder cases from the 1990s, which together had a signiﬁ cant impact on the conduct of criminal investigation in England and Wales. R v Paris, Miller and Abdullahi, 1 popularly known as the case of the ‘ Cardiff Three ’ , became both a cause cé l è bre miscarriage of justice and a leading authority on the admissibility of confessions in English law. The second case is less well known, partly because the judgment was not for- mally reported. It concerns the unsuccessful prosecution of George Heron for the murder of Nikki Allan in 1992. One way of telling the story of these cases is as a progressive strengthening of criminal justice integrity, due process values and professionalism in policing. This is an approach which, as will be shown below, has some validity. Progress in criminal justice reform should not be underestimated, not least because doing so threatens to undermine opposi- tion to contemporary assaults on integrity, due process and professionalism in criminal investigation. 2 On a broader view, however, a more nuanced and less Whiggish history may be necessary, one which pays attention to the experiences and perceptions of individuals and communities affected by injustice and which helps us to understand the criminal process as opaque and often riddled with uncertainty and the concept of integrity as capable of conﬂ icting interpretations.

1 R v Paris, Miller and Abdullahi (1993 ) 97 Cr App R 99 (CA). 2 D Dixon , ‘ Authorise and Regulate’ in E Cape and R Young (eds), Regulating Policing (Oxford , Hart Publishing, 2008 ) . 76 David Dixon

(a) The Murder of Lynette White and the Conviction of the Cardiff Three

Anthony Paris, Stephen Miller and Yusef Abdullahi were convicted of the murder of Lynette White in 1990.3 (Two other defendants, John Actie and Ronald Actie, were acquitted.) Two years later, the Court of Appeal allowed their appeals and quashed all three convictions. The Court found that the key confession made by Stephen Miller, during what Lord Chief Justice Taylor described as a ‘ travesty of an interview ’,4 had been obtained by oppression and should have been excluded from the trial. 5 The Court of Appeal was ‘ horrifi ed’ by the interrogation: ‘ Short of physical violence, it is hard to conceive of a more hostile and intimidating approach by offi cers to a suspect.’ 6 The presence of legal advisors during the interviews had evidently not protected these suspects. South Wales Police initially reacted angrily to the decision, making clear that their investigation was fi nished: as far as the police were concerned, those responsible for the murder had been caught and prosecuted. 7 The case was eventually re-opened, largely due to the persistence of an investigative journalist, Satish Sekar, 8 and in 2003 Jeffrey Gafoor was convicted of Lynette White’ s murder on the strength of new DNA evidence. Gafoor was one of the victim ’ s sex work clients. He had acted alone and had no connection whatever to Miller, Paris or Abdullahi. With the innocence of the Cardiff Three positively established, South Wales Police found themselves embroiled in a decade of further controversy surrounding the investigation. Gafoor had cut himself during the murder, leaving his blood (containing his DNA) at the scene. Heavily blood-stained with his own and his victim’ s blood, he was seen leaving the area by a witness whose description was used to produce a photofi t image. When eventually arrested, ‘ Gafoor accepted in interview that the photofi t could well have been him’ . 9 Even allowing for the wisdom of hindsight, South Wales Police had made a serious error of judgement in shifting the focus of the investigation from fi nding the man in the photofi t to pursuing the Cardiff Three.

(b) The Murder of Nikki Allan and the Trial of George Heron

In 1992, Nikki Allan, a seven-year-old girl, was murdered in a derelict Wearside warehouse after leaving her mother to walk home.10 The police identiﬁ ed George Heron as a suspect

3 See S Sekar , Fitted In : The Cardiff Three and the Lynette White Inquiry (London , The Fitted In Project, 1997 ) ; JL Williams , Bloody Valentine : A Killing in Cardiff ( London , HarperCollins , 1995 ) . 4 Paris, above n 1, ( 1993 ) 97 Cr App R 99, 104. 5 Pursuant to the Police and Criminal Evidence Act 1984 (PACE), s 76(2)(a). 6 Paris, above n 1, ( 1993 ) 97 Cr App R 99, 103 . 7 South Wales Police, Lynette Deborah White Phase III Inquiry : Final Report pursuant to Section 73(1) The Police Act 1996 South Wales Police Misconduct Reference — 35/2003 (Cardiff , South Wales Police, 2013 ) 3 . 8 See S Sekar , The Cardiff Five : Innocent Beyond Any Doubt (Hook Hants, Waterside Press, 2012 ) ; Sekar, above n 3. 9 South Wales Police, above n 7, 19. 10 My discussion of this case draws on George Heron ’ s audio-taped and transcribed interview records, Northumbria Police ’ s Report of an enquiry into the practices and procedures adopted by police offi cers during interviews with George Heron ( 1994 ) , and interviews with the principal investigating offi cer, Detective Superintendent John Renwick; Heron ’ s counsel, Mr Roger Thorn QC and Mr Robin Patton; and Dr Michael Stockdale and Greer Hogan (Northumbria Law School). I am grateful to them for their assistance, but the interpretation is mine. Integrity, Interrogation, Injustice 77 after they were given reason to believe that his statements (about his knowledge of the dead girl and his movements on the night of the murder) to offi cers carrying out house- to-house inquiries were untrue. Witness statements placed Heron in the area from which Nikki Allan disappeared. However, the crucial witness who saw the girl walking away with a man described someone unlike Heron and did not identify Heron when an identifi cation parade was held. Just as in the Cardiff Three case, investigators had evidence which did not fi t their case theory: instead of changing their theory, they tried to fi t evidence into it. As we will see, allegiance to this theory led them to misrepresent the witness’ evidence in an attempt to obtain a confession from Heron. The focus on Heron, rather than on the man described by the witness, may mean that no-one will ever be brought to justice for killing Nikki Allan. After his arrest, Heron was questioned for a total of almost eight hours in fi ve interviews over three days, during which time he denied the murder some 120 times. The fi nal interviews, in which admissions were made, were conducted by senior offi cers in the presence of a legal advisor. By the time that George Heron ’ s case came on for trial at Leeds Crown Court, the Court of Appeal’ s decision in Paris, Miller and Abdullahi had been handed down, and it clearly infl u- enced the course of the proceedings. Ruling on the voir dire, Mr Justice Mitchell declared Heron ’ s ‘ confession ’ inadmissible on grounds of oppression. In the crucial fi nal interviews, the interrogators had misrepresented the strength of the evidence against him, repeatedly asserted his guilt, asked offensive questions about his sex life, and suggested that it was in his own interests to confess. 11 Without any confession, the prosecution ’ s case collapsed and Heron was acquitted. Northumbria Police established an enquiry into ‘ the practices and procedures adopted by police offi cers during interviews’ with Heron.12 However, echoing South Wales Police, the head of Northumbria CID announced that ‘ there was no question of … “ looking for anyone else ” ’, 13 with the clear implication that the police believed George Heron was their man. Heron went into hiding (a religious order took him in). Nikki Allan’ s family subsequently brought a successful civil action for ‘ battery leading to the death of Nikki Allan’ against Heron, who did not appear to contest the case. 14 When the double jeopardy prohibition was relaxed to enable acquittals to be quashed on the basis of ‘ new and compelling evidence ’ , 15 Nikki Allan ’ s mother called for re-investigation of the case, including exhumation of her daughter ’ s body.16 Northumbria Police did reopen the investigation, leading to the arrest of another suspect (not George Heron) in 2014, but nobody was ever charged.

11 Four interview records which included the crucial admissions were excluded. Mitchell J found that misrepresentation of identiﬁ cation evidence was enough to require exclusion under PACE s 76, but insisted that the interview records should be evaluated as a whole: they constituted ‘ a continuing injustice ’ and ‘ a tale of an insidi- ous form of oppression’ . He added that even if s 76 had been found to be inapplicable, he would have excluded the recorded admissions under the s 78 fairness discretion (Mitchell J, ruling on the voir dire in R v Heron, Leeds Crown Court, 1 November 1993, 23, 26) . 12 Northumbria Police, above n 10. 13 Detective Chief Superintendent Barry Stewart, quoted in M Pithers , ‘ Nikki Allan defendant says he will sue police: Detectives “ not looking for anyone else ” ’, The Independent, 23 November 1993. 14 Northumbria Police, above n 10, 42. 15 Criminal Justice Act 2003, Part 10. 16 ‘ Pympurnell ’ , ‘ Is Justice blind— then can Nikki Allan have some? ’ , blogpost 11 January 2012, my.telegraph. co.uk/members/pympurnell/ (accessed 13 January 2016). 78 David Dixon

Although criminal procedure in England and Wales was extensively reformed by the Police and Criminal Evidence Act 1984 (PACE), changes in institutional culture and working practices took longer to disseminate and embed. The judicial interventions in Paris and Heron clearly signalled that changes in police practices and training would be needed to keep pace with statutory reforms (and that judges were prepared to play their part in infl u- encing such change through their decisions). Tom Williamson, the English police ’ s leading expert in interrogation training, reported that the Cardiff Three and Heron decisions forced ‘ a fundamental reappraisal of how the police have traditionally gone about the investigation of the most serious crimes’ .17 The Royal Commission on Criminal Justice endorsed reform of interview training, after noting that, ‘ [t]he risks inherent in certain interview techniques were vividly illustrated by the copy of the tape sent to us by the Chief Justice after the Court of Appeal had allowed the appeal of the Cardiff Three ’. 18 Most police observers could dissociate themselves from judicial criticisms of interviewing practices characterised as horrifying. In Heron’ s case, by contrast, the investigators interviewed in a way that most long-serving police offi cers would have considered unexceptionable and standard, although some Northumbria Police offi cers who were aware of developments in interview methods were critical of the senior offi cers who took over the fi nal stages of Heron ’ s interrogation. The head of Northumbria CID announced that ‘ there would be a full review of the techniques used by his offi cers’ , but added that ‘ he did not accept that these had been oppressive’ . 19 As the Northumbria Police report on the case drily observed, Mitchell J’ s ruling ‘ undoubtedly came as a surprise to many of those involved in the case’ . 20 The principal interviewers were very experienced senior detectives (rather than the detective constables involved in the Cardiff Three case); a legal advisor (albeit not a qualifi ed solicitor) had been present during audio-taped interviews without any complaint; and the tapes were ‘ vetted ’ by the Crown Prosecution Service and by a psychologist acting as an ‘ independent assessor ’ . The investigating offi cers’ superior maintained throughout that, ‘ [t]hese interviews were conducted properly by police in accordance with the Police and Criminal Evidence Act and much of the lines and styles used in the questioning of the suspect have been used over a number of years ’ . 21 However, these standard methods were clearly no longer regarded as acceptable by the criminal courts. Paris and Heron paved the way for comprehensive reform of police questioning practices through the adoption of investigative interviewing and the ‘ PEACE ’ programme. 22 Notably, this new investigative interviewing strategy extended to witnesses as well as suspects— a vital lesson learnt from the Cardiff Three investigation, as we will see. It also precipitated reform

17 T Williamson, ‘ A commentary on the ruling on voir dire ’ , in Northumbria Police, above n 10, 49. 18 Report of the Royal Commission on Criminal Justice , Cm 2263 ( London , HMSO , 1993 ) 12 – 13 . 19 Detective Chief Superintendent Barry Stewart, quoted in Pithers, above n 13. 20 Northumbria Police, above n 10, 25. 21 DCS Stewart, quoted in The Times , 2 November 1993. Also see reports and letters in The Times , 23 and 24 November 1993. 22 PEACE is a mnemonic for the investigative interviewing framework which consists of: planning and preparation; engage and explain; account; clarify and challenge; closure; and evaluation. See B Ord , G Shaw and T G r e e n, Investigative Interviewing Explained (Chatswood , NSW , LexisNexis , 2004 ) ; T Williamson (ed), Investi- gative Interviewing (Cullompton , Willan , 2006 ) ; T Shepherd and A Griffi ths, Investigative Interviewing (Oxford , Oxford University Press, 2013 ) . Integrity, Interrogation, Injustice 79 of the organisation and management of criminal investigations, with a view to neutralising the pernicious effects of prematurely narrow case theories, ‘ heroic ’ detective culture, and tunnel vision in evidence-gathering which have contributed so much to miscarriages of justice over the years. Notwithstanding strong local feelings about Heron’ s acquittal, the inter- nal inquiry was — to the credit of Northumbria Police — thorough and objective, canvassing invited submissions from professional (including academic) experts. The fi nal report was submitted to national police agencies with a clear message that change was required. The cases of the Cardiff Three and George Heron were landmarks in the modern reform of criminal investigations. We will now reconsider their broader signifi cance when re-examined through the lens of integrity. As the Introduction to this volume shows, ‘ integrity ’ has a range of meanings and potential implications for criminal proceedings. The following discussion develops the themes of integrity as a coherent and integrated set of process values; integrity as professional responsibility; integrity as truth-fi nding; and integrity as fairness. In conclusion, we will also consider a darker side to the ‘ integrity ’ of criminal process.

1. Integrity as Normative Coherence and Integration

A central meaning of integrity, for our purposes, is that important values and principles are integrated— and implemented— throughout a process, organisation or set of practices. On a holistic view, integrity implies normative coherence rather than fragmentation or irreconcilable conﬂ icts in fundamental commitments. The reality of criminal process may not be accurately depicted by the systemic ﬂ ow charts produced by administrators, but its fundamental values and commitments should be shared by its principal agencies and honoured at every stage of the process. Normative coherence presupposes effective communication between the component parts of the system. If particular parts of the criminal justice apparatus become isolated, alienated or disenfranchised, the aspiration to shared values may become more rhetorical than enacted. Regarding police interrogation, communication about what happened in the interview room used to depend on verbal or written accounts supplied by those present. Courts conventionally accepted the police version of events. Sometimes a political commitment to ‘ law and order’ was revealed, as judges lambasted counsel who dared to challenge the veracity of police accounts. 23 More prosaically, judges and magistrates were obliged to proceed on the basis of the evidence which came before them. If a confession or other evidence had been procured in accordance with accepted standards, they were generally content to work with what they had. Electronic recording of police interrogation, introduced in England and Wales in the late 1980s,24 was an innovation which had a major impact on investigative practice. When used properly, to provide a reliable record of the entire interaction between suspect and

23 C f D D i x o n, Law in Policing : Legal Regulation and Police Practices (Oxford , Clarendon Press, 1997 ) 194 – 95 . 24 Implementing PACE Code of Practice E. England and Wales have generally relied on audio-taping, in contrast to Australian jurisdictions which have used audio-visual recording since the early 1990s: see D Dixon , Interrogating Images ( Sydney , Institute of Criminology , 2007 ) . 80 David Dixon interrogators, electronic recording enables judges to overcome the practical obstacles which had formerly excluded them from the interview room, and to impose curial values and standards on police interrogators. 25 Had the courts in Paris and Heron been presented only with each suspect’ s fi nal written ‘ confession ’ , the problematic nature of their production would not have come to light. Even after Heron ’ s acquittal, the interviewers found it hard to accept that there was a problem in how they obtained his admissions. It seems certain that the construction through suggestion and adoption of key phrases in Heron ’ s confession would not have appeared in evidence had they not been electronically recorded. This is not to suggest that the police would have deliberately behaved improperly. Rather, in retrospect, the precise content of these crucial exchanges — the signifi cance of which becomes apparent only when the tape recordings were subjected to forensic scrutiny — might not have been remembered or regarded as important. Electronic recording should be understood as a potentially vital tool in securing the integrity of criminal justice, by promoting normative coherence between the interview room and the court. However, the proviso that electronic recording must cover the entire interaction, and not just a set-piece confession (as it does, for example, in many US jurisdictions 26 ), must be underlined. Incomplete recordings present a partial picture of the interview which could potentially be very misleading and damaging to the suspect (and, indeed, to the police, if selective recording leads to adverse inferences of police impropriety), and for this reason should be regarded as a threat to integrity. 27 Nonetheless, even complete tape recordings themselves guarantee nothing. One of the troubling features of the Cardiff Three’ s case was that the most oppressive sections of the interrogations were not presented to the jury which convicted them. 28 The ways in which tape recordings of interrogations are, or are not, utilised in criminal proceedings implicate the professional responsibilities of police offi cers conducting interviews, of courts reviewing confessions adduced in evidence, and— crucially, for our purposes — of lawyers providing custodial legal advice to suspects in the police station and advocates presenting the defence and prosecution cases in court.

2. Integrity as Professional Responsibility: The Role of Legal Advisors

The integrity of criminal process is partly a function of the personal integrity of its key personnel. Corrupt, malevolent or merely incompetent police ofﬁ cers pose obvious threats

25 Ibid. See also D Dixon , ‘ Video Technology and Police Interrogation ’ in GJN Bruisma and DL Weisburd (eds), Encyclopaedia of Criminology and Criminal Justice ( New York , Springer , 2014 ) . 26 RA Leo , SA Drizin , PJ Neufeld , BR Hall and A Vatner , ‘ Bringing Reliability Back In : False Confessions and Legal Safeguards in the Twenty-First Century’ [2006 ] Wisconsin Law Review 479 ; TP Sullivan , ‘ The Time has Come for Law Enforcement Recordings of Custodial Interviews, Start to Finish ’ ( 2006 ) Golden Gate Law Review 175 ; BL Garrett , ‘ Interrogation Policies ’ ( 2015 ) 49 University of Richmond Law Review 895 . 27 D Dixon , ‘ “ A Window into the Interviewing Process” ? The Audio-visual Recording of Police Interrogation in NSW Australia’ (2006 ) 16 Policing and Society 328 . 28 The trial was run twice. The fi rst trial had to be abandoned when the trial judge died as the defence was completing its fi nal submissions. The full interview record had been played in this trial. However, in the second trial, Miller ’ s barrister presented only a section of the crucial interview. His performance is strongly criticised by Sekar, above n 3, ch 8. Integrity, Interrogation, Injustice 81 to the integrity of criminal investigation. Defence lawyers also have crucial roles to play, as their failures in the Cardiff Three and Heron cases implicitly demonstrate. In an adversary system, defence lawyers are often perceived as the prosecution ’ s antagonists and may be regarded as obstructing justice. Too little attention is paid to their (real or potential) role and responsibility in giving practical substance to process values, as refl ected in section 58(1) of PACE. 29 Mention has already been made of the fact that the presence of a solicitor during their interviews had not much helped the Cardiff Three. Referring specifi cally to Stephen Miller ’ s treatment, the Court of Appeal remarked that ‘ the solicitor appears to have been gravely at fault for sitting through this travesty of an interview’ . 30 George Heron ’ s legal advisor, who was an unqualifi ed solicitor ’ s ‘ representative ’ , 31 proved no more effective. At the conclusion of the last interview with Heron, he declared himself ‘ very pleased with the way it has been dealt with’ . Ruling on the voir dire, however, Mitchell J expressed ‘ the very greatest concern’ about the fact that Heron had not seen a qualifi ed solicitor until he appeared in court, adding that ‘ [t]his is a state of affairs which must never be allowed to occur again’ . 32 Mounting concern at the failure of legal advisors to protect their clients’ interests in these and other contemporary cases prompted major reforms, including programmes of training and accreditation and disavowal of the previously widespread practice of solicitors’ fi rms using unqualifi ed, ex-police, ‘ runners ’ as legal advisors. 33 The role of defence lawyers in providing support and professional legal advice to police stations detainees needs to be understood as contributing to the integrity of criminal process by vindicating the values of an adversarial procedural system; even if it might sometimes be hard for police and prosecutors to appreciate their contribution through the din and fury of contested litigation. Sometimes, the defence lawyer’ s role is to facilitate communication between suspects and investigators, explaining the legal situation and encouraging cooperation when appropriate. On other occasions, exemplifi ed by Heron and the Cardiff Three, defence lawyers need to stand up for the law’ s fundamental principles in protecting their clients. This lesson remains to be learnt in Australia, where restriction of the right to silence in New South Wales appears to have been designed primarily to exclude lawyers from interrogation rooms. 34

29 Providing that: ‘ A person arrested and held in custody in a police station … shall be entitled, if he so requests, to consult a solicitor privately at any time. ’ 30 Paris, above n 1, ( 1993 ) 97 Cr App R 99, 104 . 31 One of the stranger details of the story is that the principal solicitor of the ﬁ rm who supplied the ‘ advisor ’ to Heron was actually in the police station at the same time, advising another suspect— who subsequently testiﬁ ed for the prosecution against Heron. 32 Mitchell J, above n 11. 33 See Dixon, above n 23, ch 6; E Cape , ‘ Assisting and Advising Defendants before Trial’ in M McConv ille and G Wilson (eds), The Handbook of the Criminal Justice Process (Oxford , Oxford University Press, 2002 ) ; L Bridges and J Hodgson , ‘ Improving Custodial Legal Advice’ [1995 ] Criminal Law Review 101 . 34 D Dixon and N Cowdery , ‘ Silence Rights ’ ( 2013 ) 17 Australian Indigenous Law Review 23 . On European developments, see J Blackstock , JS Hodgson , E Cape , T Spronken and A Ogodorova , Inside Police Custody : An Empirical Account of Suspects ’ Rights in Four Jurisdictions (London , Intersentia , 2014 ) . 82 David Dixon

3. Integrity and Truth-ﬁ nding: Themes and Accounts

The idea of integrity in police interrogation may easily be assimilated to a conception of criminal investigation as a ‘ search for the truth’ . To be sure, this is an appealing cliché : investigators conventionally say that they seek the truth, and how could anyone object to that ? However, truth and interrogation have a more complex relationship than appears at fi rst sight. All too often, investigators have already committed themselves to a ‘ truth ’ , that is an explanation of the crime and identifi cation of its perpetrator, before questioning begins. An interrogation is then less about investigation and discovery of new evidence than a concerted effort to get the suspect to accept and adopt the investigators ’ truth, their version of events. Indeed, according to the ‘ Reid Technique’ , the highly infl uential interrogation method promoted by US company John E Reid & Associates, investigators should only interrogate suspects whom they believe to be guilty. 35 The perils of the Reid Technique as an instrument of truth-fi nding are all too apparent in the transcript of George Heron ’ s interrogation. Investigators repeatedly rejected out of hand Heron ’ s attempts to establish any truth challenging their own version: A I didn’ t kill her. Q Well I believe you did kill her George. A I didn’ t. Q I believe you did. A I didn’ t … Q George you ’ ve got to help to tell the truth about this. A I am telling the truth … Q Come on, the truth George, come on we ’ re not going to mess about, you ’ ve told lies, there ’ s lies all the way along the line, you know what happened to Nikki Allan. Q You do George and you know exactly, her death … A I don’ t … Q Well let ’ s start telling the truth George, I ’ m just asking you a simple question, right, I believe I know the answer to it and I ’ m asking you to tell the truth. A I am telling the truth … A I didn’ t kill her. Q You can keep saying that over and over again but that doesn ’ t mean to me that you didn ’ t do it and you know sat there the truth … all the evidence is pointing straight in your direction … A I am not admitting to somet’ t I didn’ t do. Q We are not asking you to admit that you didn ’ t do, we are asking for the truth about the murder of Nikki Allan …

35 FE Inbau , JE Reid , JP Buckley and BC Jayne , Criminal Interrogation and Confessions, 5th edn (Burlington MA , Jones and Bartlett , 2013 ) . For criticism, see D Dixon, ‘ Questioning Suspects: A Comparative Perspective ’ (2010 ) 26 Journal of Contemporary Criminal Justice 426 . Integrity, Interrogation, Injustice 83

Q … why have you been telling lies ? Q Now George you have been asked that question a few times tonight and you have never answered it. Are you going to tell us why you have been telling lies ? It ’ s because you have got something to hide isn’ t it? … Q Well why have you been telling lies ? Why won ’ t you answer ? A How would you like to be accused of a murder that you didn ’ t commit ? … Q You might have convinced yourself you haven ’ t killed Nikki but George at the end of the day you are going to have to face up to reality you are going to have to tell the truth … Q I ’ m just asking for the truth George, that is all I am trying to establish. A I am telling the truth. In these exchanges, the police repeatedly demand their truth (acceptance of their version of events and a confession) while Heron responds by asserting an alternative truth (that he did not kill Nikki Allan). How and why do suspects making such staunch initial denials sometimes come to speak the investigators ’ truth ? As we will see, Heron eventually made an admission to knowing that Nikki Allan was dead before this became public knowledge. His initial admission opened the door to an intensive period of questioning about the circumstances of her death, and the gradual merging or imposition of the investigators’ truth into his ‘ confession ’ . Experience of miscarriages of justice in the US warns of the dangers of apparently con- vincing confessions bolstered by information fed to suspects (often unwittingly) by investigators themselves. 36 Heron conﬁ rms these dangers, but not in an entirely straightforward way. Some of George Heron’ s initial statements did not ﬁ t with information already known to police; and some of his inaccurate admissions would, if true, have made the crime even more heinous than it actually was. Interviewers responded by steering Heron away from inaccurate admissions and, at crucial moments, supplying ‘ cues ’ to help him confess ‘ accurately ’ . Thus, when Heron’ s version of the killing departed from the medical evidence, he had to be corrected and prompted: Q How many times do you think? A I lost count … Q So you have hit her with a brick a couple of times on the head ? A Yes. Q That is not all you did George is it? … Q Did that kill her George with the brick ? George it didn ’ t did it, George ? Q There is more isn ’ t there George … A Went to throttle her … Q But you hit her with something else didn’ t you? A Probably I can ’ t remember. Q George think I know it’ s not very nice but just think what else did you hit her with … A Fist.

36 R L e o, Police Interrogation and American Criminal Justice ( Cambridge MA , Harvard University Press , 2008 ) ; BL Garrett , ‘ Contaminated Confessions Revisited’ (2015 ) 101 Virginia Law Review 398 . 84 David Dixon

Q What else, howay George you used something else didn ’ t you, George, we knowhoway, George what else did you do, come on. A Piece of metal. Nikki Allan had neither been throttled nor punched. Heron had to be corrected until he eventually volunteered ‘a piece of metal ’ . But what kind of metal ? Q What did you do with it? A Hit her. Q Hit her where? A (Unintelligible). Q No you didn ’ t George, tell us what you did with it … Q But you did something else to her didn ’ t you, you say with a piece of metal, what did you do? Q Well George we know she died as a result of what happened, don ’ t we ? A Yes … Q And you’ ve told us that you’ ve hit her with a piece of metal, now you did something with that piece of metal didn ’ t you I just want you to tell us what you did … A Stuck it between her legs. Q Stuck what between her legs? A Metal pipe. Q And what did it do? A Blood … Q Where did the blood come from? From her head? A From between her legs. Q From between her legs? Howay George, look us in the eyes and tell me the truth … Q Well I’ m saying to you, that you did something else to her as well didn’ t you eh, didn’ t you George, George didn’ t you? … A Assaulted her. Q What do you mean you assaulted her? Q What did you do ? A Tried to have sex with her … Q What did you do, George, you didn ’ t try to have sex with her did you George ? A No … A I don’ t remember. Q Yes you do, right we ’ ll come back to that in a minute … Nikki Allan was not assaulted in either of the ways suggested by Heron. The investigators had to bring him back to ‘metal ’. Again, what kind of metal ? Q Now you ’ d hit Nikki when she was lying on the ﬂ oor, you hurt her again didn’ tyou George you hurt her with something … A Yes. Q What did you use? Come on. A Metal. Q A metal what ? A B a r . Integrity, Interrogation, Injustice 85

Q Bar ? A Well, a piece of metal. Q And what did you do with that piece of metal, was it a knife, George ? A It was sharp. In this crucial exchange, the investigator shifts from correcting Heron to prompting: not a bar, a knife. Well, ‘ it was sharp ’. Mention of a knife is fi rst introduced by the police, not by Heron. But in a subsequent exchange, Heron responds directly to the cue by mentioning a knife. Q … what sort of metal are we talking about? A Sharp. Q Sharp metal. A Metal. Q What are we talking about though, was it an object ? A Small, sharp, metal … Q What was this sharp metal object ? A Knife. Taken in isolation, this fi nal, brief exchange appears to record Heron admitting to using a knife to kill Nikki Allan. Viewed in the context of what came before in his interviews, however, Heron may have merely been echoing the interrogator’ s prompt that a knife was used. At any rate, the police were now evidently making progress in winning Heron over to their case theory, their truth. The murder weapon was a knife. But where did it come from ? Q It was sharp, where did you get it from. George … did you have it with you ? A N o . Q You must have had it with you. A I don ’ t remember having it with me … The investigators believed that the knife came from Heron’ s apartment, so another prompt is necessary: ‘ you must have had it with you ’ . Notably, the investigators ’ working theory begs at least two signifi cant questions. First, why would Heron be walking about with a knife from his apartment, unless the assault was premeditated? Yet premeditation was at odds with the police theory of a panicked impulsive killing. Secondly, if Heron had used the knife to kill Nikki Allan, would he really have taken it home, washed away any detectable traces of blood, and replaced it in his kitchen drawer? Yet Heron was not questioned about these matters, presumably either because the police failed to notice their signifi cance (blinkered by their attachment to a presumption of guilt) or, possibly, because they expected Heron ’ s answers to be inconvenient for their preferred account. The next issue to be resolved was how the knife was used: Q Now you have hit her with the brick, and she had another injury or injuries on her body … How did you do them ? A Wounds. Q Wounds, how did you cause the wounds? A Metal wounds … 86 David Dixon

Again, the investigators cue their suspect by indicating that the victim suffered multiple wounds: Q So we have got you, causing Nikki wounds to the torso with a knife. Do you know how many blows you would have rained on her with the knife, George ? A N o . Q Was there a lot of blows, George ? A Probably yes … Finally, where did the killer leave Nikki Allan ’ s body ? Q Was she left where you hit her? A No … Q You moved her— where to? A Near the entrance. Q Near the entrance. That is not true is it George. We know where she was found. A Basement … On numerous occasions during these exchanges, Heron evidently gives the ‘ wrong ’ answer (Nikki Allan was not throttled or sexually assaulted, or assaulted with a pole, or left near the entrance) and has to be corrected by the police. The suggestion that a knife was used originates with the police, and is later adopted by Heron. The manner in which the knife was wielded during the assault is heavily cued: asking Heron how many blows he ‘ rained down ’ on a victim and then inquiring whether there were ‘ a lot of blows ’ channelled him in the ‘ right ’ direction, though at times he seemed unsure and was possibly speculating. 37 A perplexing characteristic of the Reid Technique is its surprising indifference to ‘ the truth’ of any suspect’ s confession. Having already decided that a particular suspect committed the crime under investigation, the interrogator’ s primary objective is to ease the suspect into confessing by suggesting an account of what might have happened which the suspect is able to accept and adopt. The recommended tactic is to minimise the suspect’ s culpability or in some other way ‘ normalise ’ their offending, almost irrespective of what was actually done. The Reid Technique specifi es ‘ Nine Steps of Interrogation’ . The second step is ‘ Theme Development ’ , wherein the interrogator presents ‘ a “ moral excuse ” for the suspect ’ s commission of the offense or minimizing the moral implications of the conduct ’ . Some themes, it advises, ‘ may offer a “ crutch ” for the suspect as he moves towards a confession’ 38 by ‘ presenting reasons and excuses that will serve to psychologically (not legally) justify the suspect ’ s behavior ’ : Additionally, the interrogator minimizes the moral seriousness of the suspect ’ s criminal behavior. Blame is shifted from the suspect to some other person or set of circumstances that prompted him to commit the crime … It is highly recommended that the interrogator be prepared to present at least fi ve reasons and excuses to the suspect as to why he committed the crime and at least fi ve additional ways to minimize the suspect ’ s criminal behavior. 39

37 C f C E g a n, Murderer No More (Sydney , Allen & Unwin, 2010 ) (describing a similarly hypothetical ‘ confession ’ procured from Andrew Mallard by Western Australian police). 38 Inbau et al, above n 35, 202. 39 LC Senese , Anatomy of Interrogation Themes (Chicago , John E Reid & Associates, 2009 ) 28 – 29 . Integrity, Interrogation, Injustice 87

Regarding sexual offences, Reid & Associates provide a long list of ‘ rape themes ’ for interrogators to deploy, most of which explicitly shift blame to the victim. Investigators are advised, for example, to ‘ [b]lame the victim ’ s style of dress for leading the suspect on’ or ‘ [b]lame the victim ’ s actions … such as … rejecting the suspect’ s advances’ . 40 While such victim-blaming has been roundly condemned and rejected in modern criminal justice, Reid & Associates train interrogators to blame victims. The costs of doing so extend well beyond the immediate investigation, as Coughlin observes: By using victim-blaming stories to make rapists comfortable confessing, the police risk reinforc- ing the misogynist conventions and impulses that lead some men to rape in the ﬁ rst instance, that make victims refuse to report rapes, and that make it so difﬁ cult for the system to make charges stick for any but the most violent rapes.41 An interrogation process which diverges so far from values championed elsewhere in the criminal justice process (especially regarding sexual assault law reform) is surely one lack- ing integrity. George Heron’ s interrogators did not invite him to say that Nikki Allan was to blame for ‘ leading him on ’ . However, they did resort to a ‘ common-sense ’ account of her death, which encouraged Heron to cooperate and confess by minimising his blameworthiness. Accord- ing to the interrogators ’ proffered ‘ theme ’ , the victim must have done something (perhaps she screamed? ) which provoked an uncharacteristic, panicked response only to be expected of anyone in similar circumstances. The alternative to this everyday story of momentary human frailty, Heron was made to see, was that Nikki Allan’ s killer must be some kind of ‘ monster ’ : Q … I think she probably screamed when you took her to that place and then I think you probably panicked … You never took her there to do her any harm, you are not that type of lad are you, are you George ? … Q What was it all about ? Was it something that just went horribly wrong ? A I didn ’ t murder her. Q Was it something that just went horribly wrong ? A I didn ’ t murder her Q Why will you not address the question? A Why address the question when that would be admitting that I killed her when I didn’ t ? … A I am not that evil. Q Well … perhaps you are not that evil but perhaps because of the predicament you are in maybes through no fault of your own … you maybes might have done something when you didn ’ t know what you were doing. A I didn’ t kill her. Q You see I think possibly that it’ s never intended to be a deliberate killing, maybes one course of action was set upon, … and things got out of hand … and there was no o p t i o n …

40 Ibid 219– 20. 41 AM Coughlin , ‘ Interrogation Stories’ (2009 ) 95 Virginia Law Review 1599 . 88 David Dixon

Q George, something snapped in you that night. A I never snapped. Q We all snap at times … Q Did she scream ? A I don’ t know I didn’ t kill her. Q George did she scream? A I don’ t know I didn’ t kill her … Q Did Nikki scream or did you snap? A I didn ’ t kill Nikki. In these exchanges, in which the investigators hold fi rm to their presupposed ‘ truth ’ , the killing is normalised as what ‘ an ordinary lad’ might have done, snapping under pressure when something went ‘ horribly wrong ’ . Eventually, Heron did admit to killing Nikki Allan, adopting the ‘ I panicked’ rationalisa- tion supplied by investigators. In the decisive phase of the interrogation, Heron accedes to the offi cers ’ accusation that he knew Nikki Allan was dead when people were still searching for her. He then began to give an account which fi tted with the scream/panic theme pro- vided to him earlier in the interviews: A I don’ t know what happened. Q Right you don’ t know what happened. Right you tell me what you think happened and we will start that way … A The only thing I remember is talking to Nikki outside the pub … Q What ’ s the next thing you remember? A Being in the house watching the two children while everybody went out to search. Q But you knew she was dead, didn ’ t you George, didn ’ t you, didn ’ t you son ? Eh didn ’ t you George? George come on, come on it will be better when it is all out. Didn’ t you know she was dead, didn ’t you son ? A Yes … Q Tell [us] what happened, you have already told us that you did it because you said you knew she was dead when you went searching for her. Now come on tell us what you did and tell us what it was all about. Come on. A I panicked. This was the ‘truth ’ constructed in George Heron ’ s police interviews. In a valuable discussion, Coughlin describes how stock narratives structure interrogations and produce confessions: [T]he police confessional room is a space where the truth is produced by the interrogator ’ s strategic use of narratives that exploit popular ways of thinking about the gap between legal liability and moral culpability … [B]y using interrogations stories, interrogators actively and inescapably shape the meaning of the facts by helping suspects to embed them in a coherent narrative that coincides with our normative judgments about which acts are blameworthy and which are not. 42 As Coughlin concedes, it would be foolish to insist that interrogators should never employ narratives. Although the Reid Technique may do so in an unacceptable way, narratives are

42 Ibid 1603, 1609. Integrity, Interrogation, Injustice 89 a basic mode of human communication and ‘ the legal system invariably gains access to the meaning of facts through explanations that come in a narrative form ’. 43 Offi cials distinguish between relevant and irrelevant information by selecting from a world of facts those which make sense of what investigators see by fi tting or building an emergent account: ‘ What they want is the story that makes some of the facts comprehensible ’ .44 If narrative accounts necessarily construct our realities, what are the lessons for the integrity of custodial interrogation ? First, police should stop invoking ‘ the search for truth ’ as a peremptory way of closing down discussion about what happens in interrogation rooms. If investigators fi nd truth there, it is at least partly their own construction. Secondly, narratives do not provide a neutral structure. As Griffi n argues in relation to trials: Narrative ’ s indifference to objective facts, its invitation to readers to construct parts of the tale, and expectations it raises for their sequence, signifi cance and coherence of evidence all risk distortions in fact-fi nding … If fact-fi nders process information in terms of stories rather than logic, then they are predisposed to some misleading elements from familiar plots. 45 Unlike many fi ctionalised or stereotypical narratives, real-life events are frequently messy, confusing and inconclusive. Real violence to truth can be infl icted by attempts to ‘ order ’ stories and understandings by prematurely imposing certainty or ‘ closure ’ . Heron was just such a case. George Heron’ s acquaintance with the victim and a lie he told about his whereabouts on the night of her death apparently justifi ed police suspicion against him. But Heron ’ s reaction to questioning needs to be properly contextualised, rather than viewed through the distorting lens of an idealised rationality. As a socially marginal inhabitant of an economically deprived community, Heron might be expected to relate to the police in the estranged and mistrustful manner characteristic of such areas. 46 More- over, to say that Heron lied about his whereabouts to the offi cers conducting door-to-door inquiries is a prosecutor ’ s explanatory overlay, rather than the facts speaking for themselves. The trial judge ’ s ruling on the voir dire in Heron provides the following account: In answer to the question, ‘ Do you know the victim, yes/no’ which is printed on the form, the word ‘ yes ’ is deleted. If the defendant said he did not know the child, that was not true … Another question on the form deals with his whereabouts, the question is in this form, ‘ Obtain details of movements ’ … All that is recorded against that question is, ‘ 9.30, 7/10/92, at Wear Garth, babysitting ’. 47 On any view, the record is fragmentary. How does one go from an incomplete pro forma recording only that the respondent said he was babysitting at a particular time to an accusation that he lied about his whereabouts? Asserting that he was babysitting did not necessarily imply that Heron never went outside that night. The fact that he was supposed to have been babysitting might explain why he never mentioned going out. Admittedly, the police record appeared to be corroborated by further reports ‘ that a young man bearing a strong resemblance in appearance to the defendant ’ had been seen in local pubs around the time of Nikki Allan’ s disappearance. 48 However, this evidence creates uncertainties of its own.

43 Ibid 1622. 44 Ibid 1599. 45 L K G r i f ﬁ n, ‘ Narrative, Truth and Trial’ (2013 ) 101 Georgetown Law Journal 285, 302. 46 Northumbria Police, above n 10; B Campbell , Goliath (London , Methuen , 1993 ) . 47 See Mitchell J, above n 11. 48 I b i d . 90 David Dixon

Heron’ s arrest (after police collected these witness statements) and his acquittal (after police misrepresented the statement of a witness who had seen Nikki Allan with a man shortly before her death) both relied on apparently uncritical acceptance of identifi cation evidence. Yet the fallibility of such evidence was already well known by the 1990s, 49 and there is now a substantial body of research indicating that casual identifi cation of strangers, in particular, is prone to error and should be treated with caution. 50 Unexplained absences pose further unanswered questions. The prosecution was so reliant on Heron ’ s confession that its exclusion by the trial judge caused the case to collapse. But why was there no forensic science evidence pointing to the killer ’ s identity ? Within hours of the murder, the police had identifi ed both a crime scene and the suspect’ s resi- dence where he allegedly cleaned his clothes and the murder weapon. Yet relevant physical evidence was never mentioned during Heron’ s interviews, and its absence was not raised in the ruling on voir dire or in the subsequent police inquiry. Similarly, the lack of forensic connection between the defendants and the crime scene in the case of the Cardiff Three should have raised fundamental questions about their involvement. The messiness and uncertainty of real-life cases confl ict with the general desire for clar- ity, to know ‘ one way or the other ’ who is responsible for high-profi le crimes.51 However, the resulting pressure for prompt resolution has often caused investigators to make mis- takes. In homicide investigations especially, there are strong, media-fuelled expectations of the police to solve the case, to restore social order and security, and to provide ‘ closure ’ . Miscarriages of justice are a likely result when, under intense pressure to close a case, police offi cers fall prey to tunnel vision in actively constructing a case against a known suspect while contrary evidence is overlooked or ignored.52 The narrative process has legal consequences. Coughlin argues that when investigators offer accounts to suspects, they are negotiating moral, rather than legal, responsibility. 53 This is what Reid & Associates also claim, 54 but matters are not so straightforward. Police may be interrogating to get a confession as the basis to charge: but charge with what offence ? How the account is constructed by interrogators and which options are presented to the suspect may well determine the type and gravity of offence to which a confession is ultimately made. Non-fatal offences against the person have overlapping defi nitions and various gradations, as do sexual offences. The classifi cation of homicide as manslaugh- ter rather than murder turns largely on the killer ’ s intention. Suggestions of provocation, mental incapacity or lack of specifi c intent may each have important legal consequences. Heron ’ s interrogators were evidently adept in negotiating these issues. Having persuaded Heron to admit killing Nikki Allan by mitigating his culpability for what happened, they

49 Lord Devlin , Report to the Secretary of State for the Home Department of the Departmental Committee on Evidence of Identifi cation in Criminal Cases , HC 338 ( London , HMSO , 1976 ) . 50 E A L o f t u s, Eyewitness Testimony (Cambridge MA , Harvard University Press, 1996 ) ; GL Wells , ‘ Eyewitness Identifi cation : Systemic Reforms ’ [ 2006 ] Wisconsin Law Review 615 ; BL Cutler (ed), Reform of Eyewitness Identifi - cation Procedures (Washington DC , American Psychological Association, 2013 ) . 51 K Schulz , Being Wrong (New York, Harper Collins, 2010 ) . On the messy reality of criminal proceedings, see D Sklansky , ‘ Confi ned, Crammed and Inextricable: What The Wire Gets Right’ (2011 ) 8 Ohio State Journal of Criminal Law 473 . 52 H Rossmo , Criminal Investigative Failures ( Boca Raton FL , CRC Press , 2009 ) . 53 Coughlin, above n 41, 1645. 54 Inbau et al, above n 35, 202 – 55. Integrity, Interrogation, Injustice 91 subsequently needed to change tack, by introducing a very different account emphasising a blameworthy deliberate killing: Q … when I asked you why you killed Nikki you said you didn ’ t mean to kill her is that right? A Yes. Q Well George you hit Nikki with a brick across the head … on a number of occasions and you stabbed Nikki in the body, a number of times, when you did those actions, did you mean to kill her ? A N o . Q Well I put it to you that you must have done George. At the point of the attack upon Nikki you did … Q … That to me shows a clear intention on your part to kill her at that stage. Whatever Heron might have been led to believe he was admitting to, his interrogators were constructing an admission to murder by intentional killing. Coughlin suggests that ‘ once the suspect endorses one of the plots the cops offer, the interrogation story … itself has the potential to become the past’ . 55 Offi cial or public ‘ truths ’ are those recorded in verdicts, case reports, newspapers and ‘ true crime’ stories. But there is also another past, even if we cannot recover it: there was, for example, what really happened in the fi nal minutes of Nikki Allan ’ s short life. Although it is not always possible to access such truths, a criminal justice system built on integrity would surely seek to minimise the gap between what actually happened and the recorded past. Whilst it might be naive and idealistic to expect justice systems reliant on negotiated confessions, charges and guilty pleas56 to take truth-fi nding so seriously, a criminal process committed to integrity should embrace that as its goal.

4. Integrity as Fairness

Common law systems of criminal procedure recognise procedural ‘ fairness ’ or ‘ due process’ as an independent value extending beyond accurate fact-ﬁ nding.57 Unfair criminal procedure lacks normative integrity. A critical issue for the fairness of confessions concerns the extent to which suspects may lawfully be deceived, bullied or bribed by inducements into making admissions.58 The unfairness of the interrogation tactics to which Stephen Miller of the Cardiff Three was exposed has already been described. Deception, denigra- tion, relentless repetition and ‘ pounding ’ were similarly deployed against George Heron. From the perspective of the trial judge, deception via misrepresentation of evidence was particularly problematic.

55 Coughlin, above n 41, 1609. 56 Also see McConville and Marsh, Chapter 4 in this volume. 57 See P Roberts and A Zuckerman , Criminal Evidence, 2nd edn (Oxford , Oxford University Press, 2010 ) ch 5. 58 English and Australian courts are more vigilant than their US counterparts (see Leo, above n 36) in restrict- ing the use of evidence obtained by deception: see R v Mason (1988 ) 86 Cr App R 349 (CA); but cf R v Christou and Wright (1992 ) 95 Cr App R 264 (CA); R v Palmer; R v Gyamﬁ ; R v Cooke [2014 ] EWCA Crim 1681. 92 David Dixon

In the following exchanges, Heron ’ s interrogators suggest that a witness statement put him with Nikki Allan shortly before her death. This was not true: as noted above, the witness who reported seeing Nikki Allan in the company of a man gave a description not resembling George Heron. Q And you walked away and she walked away with you didn’ t she, now again be careful of what you ’ re saying. We have evidence, I ’ m asking for the truth, come on George … A I ’ ve been telling the truth. Q You haven’ t though George, you’ ve told lie after lie man, you know as we go on the witnesses are saying hey there we go, I saw that, I saw George, there, bang … Q George, we haven’ t even started to tell you yet about the evidence we have got against you, you know we are not trying to trip you up … Later, ruling on the voir dire, Mitchell J found that this passage constituted deliberate misrepresentation of the evidence, which was fatal to the admissibility of Heron ’ s subsequent admissions. 59 Investigators were presumably hoping that Heron would be picked out in an identifi cation parade held after the interviews. As it turned out, neither the principal ‘ witness ’ nor two others who had been in the area at the time of Nicki Allan ’ s disappearance recognized Heron. The broader signifi cance of the timing of the parade is that, if the standard was an investigative interview seeking to test a suspect’ s account against previously collected evidence, it should have been conducted before rather than after the interviews with Heron. The trial judge insisted that the identifi cation parade should have preceded the interviews; or at latest should have been held when it became clear in interviews that Heron was maintaining his claim that he did not go out on the night in question. Different elements of an investigation should be coordinated in an integrity-based system. Another well-worn tactic in the interrogation play-book is to switch to the role of ‘ good cop ’ , possibly by offering suspects inducements to confess. Investigators sought to build rapport with Heron by expressing empathy for his predicament: Q George, you have probably been a poor soul who has been crying out for some sort of love is that right ? A I always do but nobody listens. Q Well George I am prepared to listen to you here tonight. A That is not the type what I was … Heron may have needed a friend, but one across the table of a police interview room was not quite what he had in mind. More signifi cantly, the police suggested to Heron — despite his repeated denials — that it would serve his best interests to confess: Q And I think it ’ s going to be in your interest to tell the truth and to get it off your chest. A I didn’ t kill her … Q George the time has come where I think it is going to be in your interests to tell the truth. A I am telling the truth.

59 See Mitchell J, above n 11. Integrity, Interrogation, Injustice 93

Heron ’ s ‘ interest ’ , at this moment, was portrayed as being seen in a more favourable light when his case went to court: Q If we knew why you did it and what made you do it, well perhaps we could understand, but the way things stand we cannot fathom out the reason behind it … Well wouldn ’ t you feel better if you told us what it was all about ? A I didn’ t kill her. Q And what made you do it. A I didn’ t kill her. Q George, if we have to prove it all the way through the courts and you are going to con- stantly deny it what are people going to think if you ’ re convicted ? … Q … if we have got to prove it without you showing any emotion people are going to look on you as an evil monster and you ’ re not are you ? A I didn ’ t kill Nikki. Such suggestions that confession was in Heron’ s best interest drew particular criticism from the trial judge. These inducements were also fatal to the admissibility of Heron’ s subsequent admissions. The audio-record of the interviews plainly reveals the insistent pressure placed on Heron to acquiesce in and adopt the ofﬁ cers ’ case theory, their truth. The line between acceptably rigorous questioning and unacceptable harassment may be a ﬁ ne one, but the trial judge was in no doubt that investigators had crossed it, making Heron’ s interrogation ‘ oppressive ’ under PACE: What occurred during that Friday night at that police station was an exercise in breaking the defendant ’ s resolve to make no admissions. The means adopted to achieve that end meant, in effect, that regardless of the fact that his eventual confession may very well have been true, the prosecution were prevented from discharging the burden imposed on them by the two limbs of Section 76(2) [to disprove vitiating factors].60 The judge found it ‘ probable that, at some stage, the defendant began to feel the relief of unburdening himself and began to speak freely’ , presumably implying that Heron’ s admissions, or some of them, may have been true. But section 76(2) mandates the exclusion of admissions obtained by oppression ‘ notwithstanding that [they] may be true ’, and in the hardest of cases, the judge did his duty by putting the integrity of the justice process above the pressure for conviction and punishment of an accused child-killer. In the case of the Cardiff Three, where the prosecution had relied on confessions bullied out of Stephen Miller, a vulnerable and highly suggestible suspect with an IQ of 75, 61 it was even more obvious that a line had been crossed and that the Court of Appeal needed to intervene to protect the integrity of the process.

60 Mitchell J, above n 11. 61 Sekar, above n 3, 39 – 43. 94 David Dixon

Conclusion: Integrity Reconsidered

Criminal justice is conventionally discussed and evaluated in terms of some kind of balance between due process and crime control, liberty and security, police powers and suspects ’ rights, law and order. However, the balance metaphor is often criticised for its limitations and shortcomings.62 Integrity may be a superior concept, emphasising adherence to laws which are deliberately designed to protect society (encompassing both victims and suspects) directly, in terms of procedural rights and remedies, and through the longer-term benefi ts of enshrining fundamental values. Integrity may demand short-term losses for long-term gains. Thus, in Heron a confession that ‘ may very well have been true’ 63 was excluded, and the prosecution’ s case collapsed, but in the longer run the police were obliged to reform their interrogation methods, thereby becoming more effi cient both in controlling crime and in respecting suspects’ rights. As well as considering integrity as procedural fairness, the two case studies explored in this chapter have elucidated further strands of criminal process ‘ integrity ’ in terms of systemic normative coherence, professional responsibility, and commitment to truth-fi nding in criminal adjudication. The conceptual limits of integrity must also be acknowledged. Its elasticity has been demonstrated by the way this account of Heron and the Cardiff Three has fi tted com- fortably within it. However, the assumption that the primary value underlying integrity is due process has to be open to challenge. Integrity may also mean the prioritisation of the state’ s interests in protecting itself, its servants and the status quo. Perhaps criminal justice lacks due process integrity inevitably in that it is the site of confl icting values, principles and interests. 64 Take criminal justice in the US as an example. If due process was primary, it becomes hard to explain the glacial pace of responses to demonstrated miscarriages of justice. Even introducing comprehensive electronic recording of interrogations is far off, despite clear evidence of its benefi ts when used properly in comparable jurisdictions. 65 Criminal justice is structured by numerous infl uences— political, economic and commercial — and a commitment to due process is just one factor infl uencing institu- tions, processes and outcomes. The complex, competing priorities of criminal justice and a less attractive face of integrity were illustrated by legal activity following the belated identifi cation of Lynette White ’ s murderer. The conviction of Jeffrey Gafoor begged questions about the original investigation and the Cardiff Three ’ s convictions. Some of these were similar to the questions about Heron ’ s case: How were statements collected from witnesses ? How was forensic evidence used? How were the inquiries managed so that they came to focus on those charged? How did inaccurate confessions come to be made? Crucially in the Cardiff Three case, why did investigators turn from the single, blood-stained man described by a witness to the belief that Lynette White had been killed by a group? While this question is not directly answered

62 See eg Dixon, above n 23; A Ashworth and M Redmayne , The Criminal Process, 4th edn (Oxford , Oxford University Press , 2010 ) . 63 Mitchell J, above n 11. 64 A Sanders , R Young and M Burton , Criminal Justice, 4th edn (Oxford , Oxford University Press, 2010 ) . 65 Dixon, above n 23; Garrett, above n 26. Integrity, Interrogation, Injustice 95 in the offi cial report on the investigation, the consequences of the shift in the focus of the investigation are exposed. 66 Once committed to their case theory, investigators collected evidence to fi t it. In this process, an unusually high number of witnesses … changed their evidence in favour of the prosecution after the defendants had been either arrested and/or charged. Signifi cantly … none of these changes [was] as a result of new information and no reasons were given as to why these witnesses had just simply changed their accounts.67 The product was a prosecution case based on an account of Lynette White’ s murder fl awed by ‘ huge inconsistencies in the evidence ’, posing ‘ many questions about its reliability ’68 and which (not just with the benefi t of hindsight) was quite implausible. At least nine people were supposedly in the small bedroom where Lynette White died. ‘ The circumstances of the murder were such that it was highly likely that there would be forensic evidence linking those involved in the murder to the scene. There was no forensic evidence linking the original defendants … to the scene ’.69 The prosecutors ’ account was based on statements about the murder from witnesses which, Gafoor ’ s conviction showed, were false. Perjury undermines the integrity of the criminal process. The authorities began an ill- fated action in response. The South Wales Chief Constable wrote to the Cardiff Three expressing regret about their conviction and committing to investigate any offences committed during the original investigation. 70 Three major (non-police) witnesses in the Cardiff Three trial were prosecuted, convicted and jailed for perjury and perverting the course of justice. They might be better regarded as victims than as criminals: their evidence against the Cardiff Three was ‘ based on pressure, intimidation, and threats [by] police offi cers ’ . 71 The three witnesses ‘ alleged that the false accounts given to the court had arisen from their fear of and acquiescence to threats and intimidation by the police offi cers who had dealt with them ’.72 These threats included that they would be charged with the murder and in one case that the witness’ child would be taken into care. 73 In sentencing them to 18 months ’ imprisonment, the trial judge acknowledged that all three of you, vulnerable in different ways as you were, were seriously hounded, bullied, threat- ened, abused and manipulated by the police during a period of several months … as a result of which you felt compelled to agree to the false accounts that they were suggesting to you.74 While conviction was inevitable (duress not being available as a defence to such charges), the lengthy jail sentences were very harsh . Attention then shifted to the investigators. When eight former offi cers went to trial in 2011, the judge ordered disclosure of various documents to the defence. The prosecution

66 South Wales Police, above n 7. 67 Ibid 21. 68 I b i d . 69 Ibid 27. 70 I b i d 5 . 71 Ibid 11. 72 Ibid 21. 73 Ibid 26. 74 Ibid 44. 96 David Dixon reported to the judge that four of the requested documents had been destroyed, submitting that it could ‘ no longer sustain a position maintaining that the court and the defendants can have the required confi dence in the disclosure process … with all its importance to our criminal justice system ’ . No further evidence was offered, the judge was invited to direct the jury to return not guilty verdicts, and the defendants were acquitted. The planned prosecution of fi ve other offi cers was abandoned.75 In a notable example of chutzpah, one of the Cardiff Three ’ s investigators, former Chief Inspector Mouncher, had his solicitor tell the world that he had ‘ always maintained the integrity of the original investigation’ , that ‘ he leaves the court with his professional reputation intact and his character unblemished’ , and that he was considering suing for false arrest. Other legal action and inquiries rumble on, with the government refusing to set up an inquiry into the collapse of the 2011 trial, an application by the Cardiff Three for judicial review of that decision, and a report by Devon and Cornwall Police on Operation Dalecrest, an inquiry into complaints from former police offi cers about the South Wales Police investigation into the original investigation. 76 Notwithstanding the limited relevance of the missing documents (ironically, they concerned complaints against the police, principally by John Actie, one of the two defendants acquitted in the Cardiff Three trial), the integrity of the process was taken to require aban- donment of any action against the police. It may invite accusations of hypocrisy to raise concerns about strict adherence to disclosure requirements. However, the Cardiff Three, the jailed witnesses and others affected had good reason to be cynical, particularly when, six weeks later, the supposedly destroyed documents were found in the possession of South Wales Police. A rather different kind of ‘ integrity ’ was preserved as the authorities, with relief, closed the book on the Cardiff Three case. A lay concept of criminal justice integrity would include an expectation that the state should be prepared to punish its offi cials when they do wrong and to provide remedies to those wronged by them. However, the authorities appear to see such action as threatening integrity of a different sort. A notably consistent and regrettable feature of miscarriage of justice cases has been failure to bring to account those whose misconduct and criminal actions caused wrongful convictions. If the authorities have not been keen to see what Lord Denning notoriously called the ‘ appalling vista ’ of recognising that police had coerced confessions and lied in court (as in the case of the Birmingham Six), they have been even less keen to punish those involved when misconduct was beyond dispute.77 Extra-judicially, Denning went further, agreeing with an interviewer that the integrity of the criminal process and ‘ the general cause of upholding the system of justice ’ required judges to ‘ put aside ’ miscarriage of justice cases. 78 Denning was unusual only in his forthright manner: his authoritarianism was characteristic of a much more widely held and deeply entrenched

75 Independent Police Complaints Commission , South Wales Police : Destruction of speciﬁ c documents leading to the collapse of the R v Mouncher & others trial at Swansea Crown Court on 1 December 2011 (IPCC , 2013 ) 6 . 76 I b i d . 77 McIlkenny v Chief Constable of the West Midlands [ 1980 ] QB 283, 323 (CA Civ); see also Dixon, n 23 above. 78 C Mullin , Error of Judgement (Dublin , Poolbeg , 1987 ) 311 – 12 . See also E Heward , Lord Denning ( Chichester , Barry Rose, 1997 ) 223 – 30 ; C Palley , ‘ Lord Denning and Human Rights’ in JJ Jowell and JP WB McAuslan (eds), Lord Denning (London , Sweet & Maxwell, 1984 ) 304 – 307, 363– 64 ; AN Wilson , ‘ England, His England’ The Specta- tor, 18 August 1990, 10; M McConville and L Marsh , Criminal Judges (Cheltenham , Edward Elgar, 2014 ) . Integrity, Interrogation, Injustice 97 commitment to systemic integrity. It would be a bad mistake to think that integrity is a concept which is monopolised by supporters of due process. Another consistent feature of miscarriage cases has been a focus on the police rather than on other players in the process. Some comments have been made about the passive role of the defence legal advisors in the Cardiff Three and Heron cases. But what of the prosecutors and trial judges in the Cardiff Three trial, given the ‘ huge inconsistencies in the evidence’ posing ‘ many questions about its reliability’ ? 79 For prosecutors, integrity must mean exercising independent judgement and not being uncritically swept along by the police case theory. It should also be pointed out that the way to the Cardiff Three’ s successful appeal was not easy. Stephen Miller’ s initial application for leave to appeal was turned down. His lawyers ‘ advised him that he had no further prospect of winning an appeal ’ . It was only ‘ by a number of accidents ’ that he ‘ found allies and renewed his application for leave to appeal ’.80 Without these ‘ accidents ’ , the Cardiff Three might never have been acquitted, and Lynette White ’ s murderer may never have been convicted. A striking feature of these cases is the gap between ideals of legal certainty and criminal justice integrity on one side and the messy reality of life on the other. If integrity involves restoration (of order, of foundational values, of just relations) then there is little sense of restoration in the lives of those affected by the cases considered here. Some of them have suffered from systemic integrity being valued above what they would see as justice. The murders of Lynette White and Nikki Allan and the trials of the Cardiff Three and George Heron have had profound and continuing effects, not just on the criminal justice process, but also on those involved and their communities. If integrity is to be a deﬁ ning principle of criminal justice, it must not be merely a synonym for due process.

79 South Wales Police, above n 7, 21. 80 Gareth Peirce, Foreword to Sekar, above n 3, x.